ENH: benchmark the in-memory lookups on hot paths

[d-chambers]

Description

Adds micro-benchmarks for the in-memory lookups DASCore repeats on its hot paths.

The existing benchmarks are all end-to-end and disk-bound, which makes them blind at this scale. A concrete example from the free-threading series: replacing the memo on _materialize_remote_file with locking alone made resolving an already-cached remote file 12.2 µs → 35.1 µs, a 2.9x regression on every remote file open. In test_scan, which opens about 40 files, that is roughly 1 ms on a 240 ms benchmark — 0.4%, comfortably inside run-to-run noise. It was found by hand, and only because someone asked the right question.

Three regressions in that series had the same shape (a cache removed in favor of a lock), and none of them were visible to CI.

What is added

benchmarks/test_lookup_benchmarks.py, covering the lookups those paths make:

• Format resolution once every plugin is loaded: the repeat load_plugins() no-op, resolution by extension, and resolution by known format/version.
• Remote cache: resolving a remote file which is already downloaded (uses the in-memory filesystem, so no network).
• IO resources: repeat handle lookup inside one operation, and ensure_local_file on a path which is already local.
• Spool access: repeat indexing and repeat length on an in-memory spool.

Each repeats its lookup enough times to do a few milliseconds of work, so the measurement is not dominated by setup. The counts differ per benchmark because the operations differ in cost by two orders of magnitude — from about 0.05 µs for the load_plugins no-op to about 13 µs for the remote-cache resolution.

Do they actually resolve changes at this scale?

Yes — checked against a change already merged. Running this file at 1183dbcc (before #779) and at current dev:

benchmark	before #779	current dev
test_load_plugins_already_loaded	0.02 s	below the 0.005 s floor
test_yield_fiberio_by_format	0.02 s	0.01 s

Those match the 4x and 1.3x improvements measured by hand when #779 landed. CodSpeed counts instructions rather than wall time, so its resolution is better still.

Notes

• Nothing outside benchmarks/ changes.
• Add the benchmark label to run CodSpeed on a PR; that gate is unchanged.
• The remote-cache fixture sets configuration with the permanent set_config tier, matching the existing convention for class- and module-scoped benchmark fixtures.

Changelog

none

Checklist

I have (if applicable):

• referenced the GitHub issue this PR closes.
• documented the new feature with docstrings and/or appropriate doc page.
• included tests. See testing guidelines.
• added the "ready_for_review" tag once the PR is ready to be reviewed.

[coderabbitai]

Warning

Review limit reached

@d-chambers, you've reached your PR review limit, so we couldn't start this review.

Next review available in: 12 minutes

Enable usage-based reviews in Billing to review now. Otherwise, wait until the next included review is available.
You're only billed for reviews past your plan's rate limits ($0.25/file).

How can I continue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews.

How do review limits work?

CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability.

For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 63e29747-ead2-4ca5-b718-770abb82c88a

📥 Commits

Reviewing files that changed from the base of the PR and between d4a21fc and a05c53f.

📒 Files selected for processing (2)

• benchmarks/readme.md
• benchmarks/test_lookup_benchmarks.py

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch micro-benchmarks

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 100.00%. Comparing base (d4a21fc) to head (a05c53f).

Additional details and impacted files

@@            Coverage Diff            @@
##               dev      #782   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files          164       164           
  Lines        17589     17589           
=========================================
  Hits         17589     17589           

Flag	Coverage Δ
network	48.31% <ø> (ø)
unittests	100.00% <ø> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[codspeed-hq]

Merging this PR will not alter performance

✅ 65 untouched benchmarks
🆕 8 new benchmarks

Performance Changes

	Benchmark	BASE	HEAD	Efficiency
🆕	test_load_plugins_already_loaded	N/A	42.3 ms	N/A
🆕	test_yield_fiberio_by_extension	N/A	86.5 ms	N/A
🆕	test_yield_fiberio_by_format	N/A	100.7 ms	N/A
🆕	test_ensure_local_file_for_local_path	N/A	64.4 ms	N/A
🆕	test_repeat_get_resource	N/A	53.7 ms	N/A
🆕	test_ensure_local_file_already_cached	N/A	67.4 ms	N/A
🆕	test_repeat_len	N/A	172.4 ms	N/A
🆕	test_repeat_patch_access	N/A	82 ms	N/A

---

_{Comparing micro-benchmarks (a05c53f) with dev (d4a21fc)}